JP3428332B2 - Image encoding method and apparatus, and image transmission method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image encoding method and apparatus, and an image transmitting method, and more particularly to encoding and recording a video signal of a moving image on an image recording medium such as an optical disk, a magnetic disk or a magnetic tape. The present invention relates to an image encoding method and device used in a system for transmitting a video signal of a moving image through a transmission path, and an image transmission method.

[0002]

2. Description of the Related Art When image data is digitally recorded on a magneto-optical disk, a magnetic tape or the like, or transmitted through a predetermined transmission medium, the data is encoded and compressed to reduce the amount of data. I have to.

Here, the so-called MPE is used as an encoding standard.
A case where the G (Moving Picture Image CodingExperts Group) coding standard is used will be described. MPEG is discussed in ISO-IEC / JTC1 / SC2 / WG11 and proposed as a standard proposal, and motion compensation predictive coding and discrete cosine transform (DCT: DiscreteCo).
sine Transform) This is a hybrid method combined with encoding. For example, US application USP 5,1 by the applicant
Proposed in the specification and drawings of 55,593 (filing date: October 13, 1992).

In this MPEG encoding, one video sequence is a GOP (frame group),
For example, the frame is divided into 15 frames, and each frame is classified into the following three types according to the prediction method. That is, I picture: (intra-frame or intra-coded picture: Intra-coded picture), picture coded in the frame, P-picture: (forward prediction or forward prediction coded picture: Predictive-coded picture), past And an image of a frame for which motion prediction is performed from a P picture or an I picture in the future, B picture: (bidirectional prediction or bidirectionally predictive coded image: Bidirectionally predictive coded picture), motion prediction is performed from a P picture or I picture in the past and future The image of the frame.

In such a so-called MPEG compression method, an image of each frame is predicted in any one of I picture, P picture and B picture prediction modes, and a prediction error is encoded and transmitted. . Since only the prediction error is basically transmitted, the data amount can be compressed as compared with the case where the image data of each frame is transmitted as it is.

FIG. 11 shows a structural example of a GOP (group of pictures) as an image group structure containing at least one I picture. Here, the number of each frame in the GOP represents the display order.

In FIG. 11, the image signals of 15 frames from frames F0 to F14 are set as a group of pictures (GOP), which is one unit of processing. In the figure,
The frames indicated by "I", "P", and "B" represent frames encoded by the I picture, P picture, and B picture, respectively. As the frame of I picture,
The image information alone is encoded and transmitted (intra-frame or intra-encoding). On the other hand, as a P-picture frame, basically, as shown in FIG. 11A, a prediction picture is an I-picture or P-picture frame that is past in time, and the prediction residual. The difference signal is encoded and transmitted (forward prediction or forward prediction encoding). Further, as a frame of a B picture, basically, as shown in FIG. 11B, a prediction residual signal is encoded and transmitted using both reference frames temporally past and future as prediction images. Yes (bidirectional or bidirectional predictive coding).

[0008]

By the way, in a series of pictures to be transmitted, at a scene change portion where scenes such as scenes and scenes are switched, from a frame before the scene change to a frame after the scene change. Even if a picture is predicted, the prediction error becomes extremely large, and the image quality of the predicted frame deteriorates. Further, if the frame whose image quality is deteriorated is used as a predicted image and the image data of another frame is compressed, the image quality of the image data of the other frame also deteriorates.

Therefore, in the prior art, as shown in FIG. 12, an I picture is inserted immediately after a scene change and a reference frame after that is replaced with a new I picture to prevent the propagation of image quality deterioration. .

That is, in FIG. 12, when there is a scene change SC at frame F7 as shown in FIG.
Picture a of frame F8, which is the P picture immediately after that
Are changed to I pictures as shown in FIG.

However, with such a conventional scene change countermeasure, the number of I pictures increases in the entire sequence. Therefore, since the I picture transmits an image signal for one frame as it is, a large amount of information is required for transmission. Therefore, an increase in the number of I pictures leads to a reduction in compression efficiency and deterioration of the overall image quality. I was invited.

[0012]

In order to solve the above-mentioned problems, the present invention provides one of a plurality of coding modes including at least intraframe coding, forward prediction coding and bidirectional prediction coding. The input image signal is encoded in a predetermined order to generate encoded data including an intra-frame encoded image, a forward encoded image, and a bidirectional predictive encoded image, and a scene change of the input image signal within the frame. Detecting using a coded image and a forward coded image, the input image signal is coded in any of a plurality of coding modes including intraframe coding, forward predictive coding and bidirectional predictive coding. , While changing the forward predictive coded image immediately after the detected scene change to the intra-frame coded image,
At least one of the intra-frame coded images before and after the detected scene change is changed to a forward predictive coded image, and the detection is performed by setting the bit amount generated when coded with a fixed quantization size to the frame. Measurement is performed for each of the intra-coded image and the forward-coded image, and when the change in the amount of generated bits between the two forward-predicted coded images exceeds a first threshold value, or when the intra-frame coded image and the forward-coded image are forwarded. It is characterized in that a scene change is determined when a change in the amount of generated bits with respect to the directional predictive coded image is within the range of the second threshold.

Further, the present invention is characterized in that at least a part of the intra-frame coded image in the same scene between the above scene changes and the next scene change is changed to the forward prediction coded image.

Here, as the detection of the scene change, the bit amount generated when encoding is performed with a fixed quantization size is measured for each frame, and the bit amount generated between two forward prediction encoded images is measured. , The change in the amount of generated bits between the intra-frame coded image and the forward predictive coded image immediately thereafter, and the forward predictive coded image and the intra-frame coded images before and after this image. Scene change detection may be performed in accordance with the amount of change in the amount of generated bits during the period. Alternatively, the average of the brightness values in the frame may be measured for each frame, and the scene change may be detected according to the amount of change between the average of the past several frames and the value of the current frame.

If the intra-frame coded image is arranged immediately before the forward prediction coded image immediately after the scene change, the above change is not performed.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION Some preferred embodiments of an image coding method and apparatus and an image transmission method according to the present invention will be described below with reference to the drawings.

FIG. 1 is a flow chart for explaining the basic operation of the image coding method according to the present invention. In this image coding method, an input image signal is coded in a predetermined order in one of a plurality of coding modes including intraframe coding, forward predictive coding and bidirectional predictive coding. Is output.

In FIG. 1, in the first step S1, a scene change of the input image signal is detected. In the next step S2, the forward prediction coded image (P picture) immediately after the detected scene change is changed to the intra-frame coded image (I picture). Further, in this step S2, at least one of the intra-frame coded images (I pictures) before and after the detected scene change is forward-predicted coded image (P picture).
Change to. That is, if there is a scene change,
Change past or future I pictures or both to P pictures, and change the reference frame after the scene change to I pictures.
It has been changed to a picture.

A specific operation at this time will be described with reference to FIGS. FIG. 2A shows, for example, two GOPs (groups of pictures) as described with reference to FIG. 11, and shows a case where a scene change SC is detected in the B picture of the frame F7. .
At this time, the P picture which is the reference frame immediately after the detected scene change SC, that is, the picture a of the frame F8 is changed to the I picture as shown in FIG. The picture b of the frame F2 which is a past I picture is
Change to P picture as shown in FIG.

The scene change SC shown in FIG. 2 (A)
On the other hand, instead of the picture b which is the past I picture, the picture c which is the future I picture may be changed to the P picture, and the frame sequence at this time is shown in FIG.
(A) of. In the example of FIG. 3A, a picture a which is a P picture immediately after the scene change SC of FIG. 2A is changed to an I picture, and a picture which is a future I picture with respect to the scene change SC. The c is changed to the picture P.

As yet another example, both the pictures b and c that are past and future I pictures for the scene change SC of FIG. 2A are P pictures as shown in FIG. 3B. You may change to. That is, in the example of FIG. 3B, the picture a which is the P picture immediately after the scene change SC of FIG. 2A is changed to the I picture, and both the past and future of the scene change SC are changed. The pictures b and c, which are the I pictures of, are changed to the picture P.

If an I picture precedes a P picture immediately after a scene change, it is not necessary to make the above change.

By changing the picture type in this way, it is possible to prevent the propagation of the image quality deterioration by arranging the I picture immediately after the scene change, and since the ratio of the I picture as a whole does not increase, the compression efficiency is improved. Can also be prevented.

Next, an example of the image coding apparatus according to the embodiment of the present invention will be described with reference to FIG.

The image encoding apparatus shown in FIG. 4 encodes an input video signal to generate first encoded data.
Encoding circuit 10, an image analysis or in-frame information circuit 60 for obtaining statistical properties or image characteristics of the input video signal, and the first encoded data from the first encoding circuit 10 at predetermined time intervals. A scene change detection circuit 3 for detecting a scene change from the amount of data and the statistical data or image characteristic information from the in-frame information analysis circuit 60.
0, a picture type designating circuit 99 for deciding an image coding method at every predetermined time based on the information from the scene change detection circuit 30, and the input video based on the information from the picture type designating circuit 99. Second encoding circuit 4 for encoding the signal to generate second encoded data
With 0 and.

The image analysis or intra-frame information analysis circuit 60 calculates, for example, statistical information of brightness and flatness, and calculates it as image characteristic information of the input image.

Specific examples of the image characteristic information will be given. As the statistical information on the brightness of the input image, for example, an average value (L) of the brightness signal Y for each predetermined time can be mentioned. The statistical information about the chromaticity of the input image may be, for example, the average value (R) of the chromaticity signal Cr for each predetermined time.
Can be mentioned. Further, as the statistical information of the flatness of the input image, for example, a variance value (V) of the luminance signal Y for each predetermined time can be cited. As the statistical information of the motion amount of the input image, for example, a motion vector amount The average value (M) for each predetermined time can be listed.

The first encoding circuit 10 stores the input image data, which is an input video signal, in the frame memory group 12, as shown in FIG. Then, the information indicating the picture type from the picture type designation circuit 100 is supplied to the motion vector detection circuit 11, and the motion vector detection circuit 11
Detects the motion vector of the input image data in macroblock units based on the image data stored in the frame memory group 12 and the information indicating the picture type. Further, the first encoding circuit 10 has a frame memory 22 for storing the predicted image data, and the frame memory 2
The predictive image data read from No. 2 is the motion compensation circuit 23.
And the motion compensation is performed on the predicted image data based on the motion vector from the motion vector detection circuit 11. Further, the absolute value sum of the inter-frame difference in macroblock units from the motion vector detection circuit 11 is supplied to the intra-frame / forward / backward / bidirectional prediction determination circuit 13, and further, in-frame / forward / backward / bidirectional The prediction determination circuit 13 determines the prediction mode based on the sum of absolute values of the supplied inter-frame differences, and controls the prediction encoding circuit 14 to switch intra-frame / forward / backward / bidirectional prediction in block units. Then, the determined prediction mode is supplied to the predictive coding circuit 14.

The predictive coding circuit 14 receives the motion-compensated predicted image data and the input image data from the motion compensation circuit 23, and inputs the predicted image data based on the prediction mode from the prediction judgment circuit 13. The difference from the image data is calculated and the prediction code data (difference data) is output.
Difference data which is a prediction error from the predictive coding circuit 14 is, for example, a discrete cosine transform (hereinafter DCT: Discrete Cos).
ine Transform) circuit, DCT circuit 1
5 generates coefficient data. The coefficient data from the DCT circuit 15 is supplied to the quantization circuit 16. The quantization circuit 16 quantizes the coefficient data with a constant quantization step size to generate quantized data, and the quantized data is quantized by a variable length coding circuit (VLC: Variable Length Cod).
e) 17 and the inverse quantization circuit 18 are supplied. The variable length coding circuit 17 performs variable length coding on the quantized data and outputs the variable length coded data to the counter. The inverse quantization circuit 18 inversely quantizes the quantized data from the quantization circuit 16 to restore coefficient data, and the coefficient data is inverse discrete cosine transform (hereinafter referred to as IDCT: Inverse).
Discrete Cosine Transform) circuit 20 is supplied. The IDCT circuit 20 decodes the coefficient data into difference data and supplies the difference data to the addition circuit 21. The adder circuit 21 receives the difference data from the IDCT circuit 20 and the motion-compensated predicted image data from the motion compensation circuit 23, adds the difference data and the predicted image data, and predicts the next input image data. Image data is generated and the predicted image data is stored in the frame memory 22.
Supply to.

The second encoding circuit 40 is similar to that shown in FIG.
, The delay device 43 delays the input image data.
Then, the input image data which is the input video signal from the delay circuit 43 is stored in the frame memory group 102. Then, the information indicating the picture type from the picture type designation circuit 99 is supplied to the motion vector detection circuit 101, and the motion vector detection circuit 101 outputs the information indicating the picture data and the picture type stored in the frame memory group 102. Based on this, the motion vector of the input image data is detected in macroblock units. Furthermore, the second encoding circuit 40
Has a frame memory 52 for storing predicted image data, the predicted image data read from the frame memory 52 is supplied to the motion compensation circuit 53, and based on the motion vector from the motion vector detection circuit 101, Motion compensation is applied to the predicted image data. Also, the sum of absolute values of inter-frame differences in macroblock units from the motion vector detection circuit 101 is supplied to the intra-frame / forward / backward / bidirectional prediction determination circuit 103, and further, this intraframe / forward / backward / bidirectional Prediction determination circuit 103
The prediction encoding circuit 4 determines the prediction mode based on the sum of absolute values of the supplied inter-frame differences, and switches intra-frame / forward / backward / bidirectional prediction in block units.
4, the determined prediction mode is supplied to the predictive coding circuit 44.

The predictive coding circuit 44 receives the motion-compensated predictive image data and the input image data from the motion compensating circuit 53, and based on the predictive mode from the predictive coding circuit 44, predictive image data The difference from the input image data is calculated and the prediction code data (difference data) is output. Difference data which is a prediction error from the predictive coding circuit 44 is, for example, a discrete cosine transform (hereinafter DCT: Discrete).
Cosine Transform) circuit 45, DC
The T circuit 45 generates coefficient data. DCT circuit 45
The coefficient data from is supplied to the quantization circuit 46. The quantization circuit 46 quantizes the coefficient data with a fixed quantization step size to generate quantized data, and the quantized data is quantized by a variable length coding circuit (VLC: Variable Len).
gth Code) 47 and inverse quantization circuit 48. The variable-length coding circuit 47 receives the quantized data from the quantization circuit 45 and the motion vector from the motion vector detection circuit and the prediction mode determined by the intra-frame / forward / backward / bidirectional prediction determination circuit 103. , And variable-length code these data and output them to the transmission buffer 49 as variable-length coded data. In addition, the inverse quantization circuit 48
Dequantizes the quantized data from the quantization circuit 46 and restores it to coefficient data, and the coefficient data is converted to, for example, an inverse discrete cosine transform (hereinafter referred to as IDCT).
It is supplied to the circuit 50 (called e Transform). The IDCT circuit 50 decodes the coefficient data into difference data and supplies the difference data to the addition circuit 51. The adder circuit 51
The difference data from the IDCT circuit 50 and the motion-compensated predicted image data from the motion compensation circuit 53 are received, and the difference data and the predicted image data are added to generate predicted image data for the next input image data. , And supplies the predicted image data to the frame memory 52. Further, the transmission buffer 49 temporarily stores the variable length code data from the variable length encoding circuit 47 and outputs it as transmission data at a constant bit rate. The transmission buffer 49 also feeds back information indicating the storage state of the buffer to the quantization scale setting circuit 33. The quantization scale setting circuit 33 controls the quantization circuit 46 according to the assigned code amount, the generated code amount, and the feedback information. Also,
The output data is recorded on a recording medium 120 such as a disk, a tape, or a semiconductor memory, or is transmitted to the receiving side via the transmission path 110. Further, the transmission data (encoded data) recorded on this recording medium is decoded by an image decoding device (not shown) and restored to the original image data.

In this image coding apparatus, the input image data from the motion vector detecting circuit 11 is supplied to the in-frame information analyzing circuit 60, and the in-frame information analyzing circuit 6
0 obtains statistical information about the input image data for each predetermined time. Then, the statistical information is supplied to the scene change detection circuit 30. Scene change detection circuit 30
Is information from the counter, picture type designation circuit 10
A scene change is detected based on the information indicating the picture type from 0 and the input statistical information, and the information indicating the detection result is supplied to the picture type designation circuit 99.
The picture type designation circuit 99 supplies information indicating the picture type to the motion vector circuit 101 based on the information from the picture type designation circuit 100 and the detection information from the scene change detection circuit 30.

Next, this embodiment will be described in more detail. The first encoding circuit 10 performs an encoding process on one sequence of input image data. That is, for example, by using the information indicating the picture type from the picture type specifying circuit 100, for example, the picture type is cyclically specified as shown in FIG. 5, and the predictive coding process, the DCT conversion process, and the fixed quantization step size are used. The quantization process and the variable length coding process are sequentially executed. Further, the scene change detection circuit 30 detects the scene change and sends the obtained scene change information to the picture type designation circuit.
Supply to 9. Then, the second encoding circuit 40 performs encoding so that the schedule of the picture type is optimal based on the information indicating the picture type from the picture type designating circuit 99. Here, the one sequence is
For example, all frames recorded on one image recording medium, such as one movie or program, are assumed.
When the recording medium is divided and used, all frames recorded in each divided area may be one sequence. Also,
The scene change detection circuit 30 uses the counter 31
A data amount of variable-length code data, which is the first bit stream obtained by the first encoding circuit, for each predetermined time;
Information indicating the picture type from the picture type designation circuit 100, and the average value (L) and variance (V) of the luminance signal Y of the input video signal obtained by the intra-frame information analysis circuit 60 for each predetermined time, the chromaticity signal. A scene change is detected based on the average value (R) of Cr for each predetermined time and the average value (M) of the motion vector amount for each predetermined time, and the picture type is specified for each predetermined time for one sequence. There is. Further, the second encoding circuit 40 again performs predictive encoding processing, DCT conversion processing, quantization processing, variable length encoding processing on the input image data based on the information indicating the picture type from the picture type designating circuit 99. The second
Variable-length code data, which is a bit stream of. At that time, the second encoding circuit 40 quantizes the input image data with a quantization step size based on the encoding bit rate.

Here, in the above-mentioned so-called MPEG or the like, a certain number of pictures are encoded as a group called a GOP. First encoding circuit 10 constituting the image encoding device of the first embodiment
A specific example of the operation will be described with reference to the case where a picture is designated in the GOP cycle in FIG.

The I picture is coded only by the picture, the P picture is coded by the I frame or the P picture which is the previous reference frame, and the B picture is created by the reference frames before and after the I picture. Therefore, if the input image and the picture type are specified in the order of each of the frames F0 to F14 in FIG. 5, the frames are encoded in the order shown in the encoding order in FIG. That is, the input frames F2, F0, F1, F5, F3, F4, F8, F6
F7, F11, F9, F10, F14, F12, F13
Are encoded in this order.

The motion vector detection circuit obtains the input image data and the picture type information, determines when the input image is coded, and when the coded time comes, the motion vector detection circuit retrieves it from the frame memory 12 and moves it. Derive a vector.

The quantizing circuit 16 of the first encoding circuit 10 shown in FIG. 4 has a quantization step size of 1, for example, D
By quantizing the coefficient data supplied from the CT circuit 15,
Generate quantized data. Counter 3 of encoding control circuit
1 indicates the generated code amount (y) by counting the data amount of the variable length code data (first bit stream) obtained by variable length coding this quantized data for a predetermined unit time, for example, for each frame. Calculate for each frame. Further, the image analysis or in-frame information analysis circuit 60 obtains the average value (L) and variance (V) of the luminance signal Y and the average value (R) of the chromaticity signal Cr in the frame, and the macro block The average value (M) of the motion vector amounts of is calculated.

The scene change detection circuit 30 measures the generated code amount (y) in the first encoding circuit 10 and considers which picture type the frame is encoded in. Observe the state of the difference.
Then, the difference amount between frames is compared with a predetermined value that is set, and the presence or absence of a scene change is determined based on the comparison result. Specifically, when the difference value exceeds the set threshold value, it is determined that the frame has a scene change.

Here, some examples of specific scene change detection methods will be described with reference to FIGS. 6 to 9.

In the specific example of FIG. 6, a scene change is detected in accordance with the change amount or difference of the generated bit amount (y) between two P pictures by using the change of the generated bit amount (y) for each picture. Are doing. That is, the current picture is a P picture, for example, the frame F8 in FIG.
, The generated bit amount (y) in this frame F8 is the P picture before this frame F in FIG.
The difference with the generated bit amount (y) in 5 is calculated. If the difference Ya exceeds a predetermined threshold value, it is assumed that a scene change has occurred between these P pictures, that is, between frames F5 to F8 in FIG. 6, and at the current P picture, frame F8. It is said that the scene change SC was discovered.

In the specific example of FIG. 7, scene change detection is performed according to a change amount or a difference between the generated bit amount (y) in the I picture and the generated bit amount (y) in the subsequent P picture. Is going. That is, when the current picture is the P picture (frame F5 in FIG. 7) immediately after the I picture, the amount of generated bits (y) in this P picture and the previous I picture (frame F2 in FIG. 7)
The difference with the amount of generated bits (y) in (1) is calculated. When the difference Yb is within the range of the predetermined threshold value, the frames F2 to F in FIG.
If there is a scene change by 5
It is assumed that a scene change is found in frame F5 which is a picture.

FIG. 8 shows a specific example in which the scene change is detected by obtaining the change amount or difference between the P picture and the I pictures before and after the generated bit amount (y). That is, when the current picture is a P picture, for example, the frame F8 in FIG.
Generated bit amount (y) in 8 and I before and after this p-picture
The difference between the picture and the frame F2 or F17 in FIG. 8 is calculated. If the difference Yc or Yd is within a predetermined threshold range, it is considered that there is a scene change between P pictures and that a scene change is found in the current P picture.

Next, FIG. 9 shows an in-frame information analysis circuit 60.
An example is shown in which the scene change detection is performed using the average (L) of the brightness values of the screen, which is the information from. That is,
When the current picture is frame F13 and the difference Le from the average value Lav of the brightness values of a certain number of frames before this exceeds a predetermined threshold value, there is a scene change in frame F13 which is that picture. As a result, a scene change was found in the current picture.

The picture type designation circuit 99 designates a picture for the second coding based on the information from the scene change detection circuit 30 as described above.

As described above with reference to FIG. 2B, whether or not the reference frame immediately after the scene change is changed to the I picture and the I picture immediately before that is changed to the P picture is specified. Alternatively, as described with reference to FIG. 3A, the reference frame immediately after the scene change is changed to the I picture, and the I picture immediately after that is changed to the P picture. Alternatively, as described with reference to FIG. 3C, the reference frame immediately after the scene change may be changed to an I picture, and both of the preceding and following I pictures may be changed to P pictures. Good. In addition, P immediately after the scene change
If there is an I picture before the picture, no change is made.

Next, the operation of the second encoding circuit 40 will be described. In addition, the quantization scale setting circuit 33, the delay device 43, and the quantization circuit 4 which constitute the second encoding circuit 40.
6. The circuits other than the transmission buffer memory 49 perform the same operations as the circuits forming the first encoding circuit 10 described above, and thus the description thereof will be omitted.

The delay device 43 delays the input image data by the time until the picture type signal is output from the picture type designation circuit 99, for example. Then, in the predictive coding circuit 44 and the DCT circuit 45, the delayed input image data is subjected to predictive coding processing according to the prediction mode supplied from the intra-frame / forward / backward / bidirectional prediction determination circuit 13, The DCT circuit 45 generates coefficient data, and the coefficient data is input to the quantization circuit 46.

When the buffer feedback information from the transmission buffer 49 indicates that the overflow of the transmission buffer 49 is near, the quantization scale setting circuit 33 does not depend on the comparison result of the above-mentioned assigned code amount and generated code amount, Increase the quantization step size to suppress overflow. Further, when the buffer feedback information from the transmission buffer indicates that the underflow of the transmission buffer 49 is close, the quantization step size is reduced and the underflow is reduced regardless of the comparison result of the assigned code amount and the generated code amount. The flow may be suppressed.

In the conventional scene change countermeasure, the number of I-pictures increases in the entire sequence, and the I-picture transmits an image signal for one frame as it is. Therefore, a large amount of information is required for transmission. The increase in the number of I-pictures leads to a reduction in compression efficiency and a deterioration in image quality. However, according to the embodiment of the present invention as described above, deterioration due to a scene change can be prevented and the image quality can be improved without increasing the number of I-pictures as a whole, and the GOP can be changed only in a part of the whole. The influence on bit allocation such as rate is small.

Next, another embodiment according to the present invention will be described. For example, there are scene changes SC1 and SC2 as shown in FIG.
In the case of being divided into N2 and SN3, the above-mentioned picture type designation circuit 99 changes the first P picture (frame F5 or frame F20) after a scene change to an I picture. Then, further, until the next scene change is detected, the I picture (frame F8,
By changing F17 or frame F26) to P pictures, the number of I pictures may be reduced as shown in FIG. 10B, and the number of I pictures in one scene may be reduced as much as possible.

As described above, by reducing the I pictures in the same scene, the compression efficiency can be improved and the image quality can be improved.

The present invention is not limited to the above-described embodiments, but the picture type configuration and the number of frames in, for example, a GOP (group of pictures) are not limited to the illustrated example, and other than that, It goes without saying that various modifications can be made without departing from the scope of the present invention.

[0053]

According to the present invention, an input image signal is coded in a predetermined order in any one of a plurality of coding modes including at least intraframe coding, forward predictive coding and bidirectional predictive coding. When generating coded data including an intra-frame coded image, a forward-direction coded image, and a bidirectional predictive-coded image, a scene change of an input image signal is used by using the intra-frame coded image and the forward-coded image. Detected,
The input image signal is coded in any of a plurality of coding modes including intraframe coding, forward predictive coding and bidirectional predictive coding, and a forward predictive coded image immediately after the detected scene change is generated. While changing to an intra-frame coded image, at least one of the intra-frame coded images before and after the detected scene change is changed to a forward predictive coded image, and the detection is coded with a fixed quantization size. When the amount of bits generated at the time of encoding is measured for each of the intra-frame coded image and the forward direction coded image, and the amount of change in the amount of generated bits between the two forward predictive coded images exceeds the first threshold value. Alternatively, when the change in the amount of generated bits between the intra-frame coded image and the forward prediction coded image is within the range of the second threshold value, it is determined that the scene change occurs. And at the same time prevent the deterioration of image quality due to di-,
It is possible to prevent an increase in the amount of transmission information by suppressing an increase in an intra-frame coded image having a large coded bit amount, improve the image quality, and have a small influence on bit allocation such as a variable rate.

By reducing the number of intra-frame coded images in the same scene, the compression efficiency can be increased and the image quality can be improved.

[Brief description of drawings]

FIG. 1 is a flowchart for explaining a basic operation of an image encoding method according to the present invention.

FIG. 2 is a diagram illustrating an example of changing a picture type according to a scene change.

FIG. 3 is a diagram for explaining another example of changing a picture type according to a scene change.

FIG. 4 is a block diagram showing an example of an image encoding device according to an embodiment of the present invention.

FIG. 5 is a diagram showing an order of specifying picture types and an order of encoding in a group of pictures.

FIG. 6 is a diagram for explaining a first specific example of scene change detection.

FIG. 7 is a diagram for explaining a second specific example of scene change detection.

FIG. 8 is a diagram for explaining a third specific example of scene change detection.

FIG. 9 is a diagram for explaining a fourth specific example of scene change detection.

FIG. 10 is a diagram for explaining a specific example of reducing I pictures in the same scene.

[Fig. 11] Fig. 11 is a diagram illustrating an example of a group of pictures and a picture type when image data is compression-encoded.

[Fig. 12] Fig. 12 is a diagram for explaining an example of conventional picture type change according to a scene change.

[Explanation of symbols]

10 First Coding Circuit, 13, 103 Intra-frame / Forward / Backward / Bidirectional Prediction Determination Circuit, 14,44 Predictive Coding Circuit, 30 Scene Change Detection Circuit, 4
0 second encoding circuit, 99 picture type designation circuit

─────────────────────────────────────────────────── ─── Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04N ^7/ 24-7/68

Claims (6)

(57) [Claims] 1. An intra-frame coded image in which an input image signal is coded in a predetermined order in any one of a plurality of coding modes including at least intra-frame coding, forward predictive coding and bidirectional predictive coding. , Forward-coded images and both
In an image coding method for generating coded data including a predictive coded image , a scene change of an input image signal is converted into an intraframe coded image.
And a detection step of detecting using a forward encoded image, and encoding the input image signal in any of a plurality of encoding modes including intraframe encoding , forward predictive encoding, and bidirectional predictive encoding. And the forward predictive coded image immediately after the detected scene change is changed to an intra-frame coded image, and at least one of the intra-frame coded images before and after the detected scene change is forward-coded. possess a changing step of changing the predictive coded picture, said detecting step, encoded with a fixed quantization size
The amount of bits sometimes generated is determined by
Two forward-prediction coded images measured for each direction-coded image
If the change in the amount of generated bits between images exceeds the first threshold value
Or the intra-frame coded image and forward prediction code
The change in the amount of generated bits with respect to the digitized image is the second threshold range.
An image coding method characterized by determining that a scene change occurs when the area is in a circle . 2. The changing step changes at least a part of intra-frame coded images in the same scene from the scene change to the next scene change to forward predictive coded images. The image encoding method according to claim 1. 3. An intra-frame coded image in which an input image signal is coded in a predetermined order in one of a plurality of coding modes including intra-frame coding, forward predictive coding and bidirectional predictive coding , Forward coded image and bidirectional predictive coding
In the image encoding apparatus for generating encoded data including an image, the intra-frame coded picture scene change of an input image signal
And detecting means for detecting by using a forward coded image, coded in any of a plurality of coding modes including the input intra-frame coded image signal, the forward predictive coding and bidirectional predictive coding And a forward predictive coded image immediately after the detected scene change is changed to an intraframe coded image, and at least one of the intraframe coded images before and after the detected scene change is forward predicted. Change means to change to encoded image
Then, the detection means coded with a fixed quantization size.
The amount of bits sometimes generated is determined by
Two forward-prediction coded images measured for each direction-coded image
If the change in the amount of generated bits between images exceeds the first threshold value
Or the intra-frame coded image and forward prediction code
The change in the amount of generated bits with respect to the digitized image is the second threshold range.
An image coding apparatus characterized in that when it is in a surrounding area, it is judged as a scene change . 4. The change means is the scene change.
From the same scene to the next scene change
At least a part of the intra-coded image
The image code according to claim 3, wherein the image code is changed to an image.
Device. 5. An intra-frame coded image is obtained by coding an input image signal in a predetermined order in any one of a plurality of coding modes including at least intra-frame coding, forward predictive coding and bidirectional predictive coding. , Forward-coded images and both
In the image transmission method of generating coded data including a predictive coded image and transmitting the coded data, a scene change of an input image signal is detected as an intra-frame coded image.
And forward coded images are used for detection and
Code with fixed quantization size
Intra-frame coded image
And two forward prediction codes measured for each forward coded image.
The amount of change in the amount of generated bits between encoded images exceeds the first threshold
Or intraframe coded image and forward prediction
The amount of change in the amount of generated bits with respect to the encoded image is the second threshold value.
When it is within the range, it is judged as a scene change, and the input image signal is generated by any one of a plurality of coding modes including intraframe coding , forward prediction coding and bidirectional prediction coding. Then, the forward predictive coded image immediately after the detected scene change is changed to the intraframe coded image, and at least one of the intraframe coded images before and after the detected scene change is forward predictive coded image. To generate modified coded data, and transmit the modified coded data as transmission data. 6. The above-mentioned change is next from the scene change.
In the frame in the same scene until the scene change of
At least part of the coded image becomes a forward predictive coded image.
Image transmission side according to claim 5, wherein the changing
Law.